An Application of a Particle Filter to Bayesian Multiple Sound Source Tracking with Audio and Video Information Fusion
نویسندگان
چکیده
Abstract – A particle filter is applied to the problem of detecting and tracking multiple sound sources by Bayesian inference using combined audio and video information. The problem is formulated within a general framework of Bayesian hidden variable sequence estimation by fusing observed information. The particle filter is then introduced as an approximation of Bayesian inference. Experiments using real-world data demonstrate that the proposed method works well in ordinary environments such as a meeting room. The computational cost of estimation is reduced significantly compared to exact Bayesian inference, while maintaining the quality of estimation.
منابع مشابه
An Efficient Target Tracking Algorithm Based on Particle Filter and Genetic Algorithm
In this paper, we propose an efficient hybrid Particle Filter (PF) algorithm for video tracking by employing a genetic algorithm to solve the sample impoverishment problem. In the presented method, the object to be tracked is selected by a rectangular window inside which a few numbers of particles are scattered. The particles’ weights are calculated based on the similarity between feature vecto...
متن کاملDetection and separation of speech segment using audio and video information fusion
In this paper, a method of detecting and separating speech events in a multiple-sound-source condition using audio and video information is proposed. For detecting speech events, sound localization using a microphone array and human tracking by stereo vision is combined by a Bayesian network. From the inference results of the Bayesian network, the information on the time and location of speech ...
متن کاملA New Modified Particle Filter With Application in Target Tracking
The particle filter (PF) is a novel technique that has sufficiently good estimation results for the nonlinear/non-Gaussian systems. However, PF is inconsistent that caused mainly by loss of particle diversity in resampling step and unknown a priori knowledge of the noise statistics. This paper introduces a new modified particle filter called adaptive unscented particle filter (AUPF) to overcome th...
متن کاملSpeaker Tracking Using Particle Filter Sensor Fusion
in Proc. of Asian Conference on Computer Vision (ACCV), 2004 Sensor fusion for object tracking has become an active research direction during the past few years. But how to do it in a robust and principled way is still an open problem. In this paper, we propose a new fusion framework that combines both the bottom-up and top-down approaches to probabilistically fuse multiple sensing modalities. ...
متن کاملA multimedia content modeling and classification methodology using visual information for the protection of sensitive user groups
The thesis concerns the problems of visual tracking and violence detection in video sequences. For the visual tracking problem, two feature fusion frameworks are presented. For violence detection, a system that classifies movie segments as violent or non-violent is proposed. The first tracking framework called ’Model Fusion via Proposal’ (MFP) framework, provides a way to efficiently fuse visua...
متن کامل